I focus on creating value. My favored tools are clear thinking, effective communication, and clean code.
I tinker with Ruby, Rust, and React Native in my downtime.
I worked remotely with the Social Updates team - mostly feature work and domain design in Ruby on Rails with Sidekiq, Redis, MySQL and Kafka, with occasional plumbing/debugging excursions into GraphQL and Vue.
Some highlights from my work here:
•
I improved an algorithm for reaggregating notifications, which decreased that endpoint's memory allocations by over 90%.
•
I built a pipeline for disseminating 1K+ peer recommendations per month on top of Sidekiq, Redis, and Kafka.
•
I wrote a backfill task to safely and efficiently update 180M+ records in our production database.
•
I optimized a daily-run archive job by making use of batching, caching, and a compound index, which reduced the job's execution time by 98%.
I use Docker containers for local development, GitHub for code reviews, and Pivotal Tracker for issue tracking.
I built tools to assist researchers and administrators with a wide variety of workflows. One large project I worked on was a Rails API/React UI microservice for visualizing respirometry data.
Some highlights from my work here:
•
I wrote rake tasks to import mission-critical Excel data for reporting to the U.S. Department of Energy, the institute's primary funding source.
•
I audited our Rspec + Capybara test suite, making fixes to align it with industry best practices, which reduced its single-process runtime by over 20%.
•
I built dynamic nested forms and a file tree navigator using AJAX + jQuery.
•
I wrote shell scripts to assist with database management and config file updates, making context switches faster and less error-prone for our entire team.
I developed against a Docker clone of a production Oracle database, using Gitlab for code reviews and Jira for issue tracking.
I worked on an import script for JPW Industries. In particular, I wrote code to alert members of the core maintenance team when parts of the import script raised errors.
•
I built the website for Vilas & Company (www.vilasandcompany.com) using Jekyll and Amazon S3.
